Abstract
BACKGROUND:
In 2008, specialty medications accounted for 15.1% of total pharmacy benefit medication spending, and per member expenditures have increased by 11.1% annually from 2004 to 2008 within a commercially insured population of 8 million members. Insurers face increasing pressure to control specialty medication expenditures and to rely on increasing member cost share through creation of a fourth copayment tier within the incentive-based formulary pharmacy benefit system. Data are needed on the influence that member out-of-pocket (OOP) expense may have on prescription abandonment (defined as the patient never actually taking possession of the medication despite evidence of a written prescription generated by a prescriber).
OBJECTIVES:
To explore the relationship between prescription abandonment and OOP expense among individuals newly initiating high-cost medication therapy with a tumor necrosis factor (TNF) blocker or multiple sclerosis (MS) biologic agent.
METHODS:
This observational cross-sectional study queried a Midwestern and southern U.S. database of 13,172,480 commercially insured individuals to find members with a pharmacy benefit-adjudicated claim for a TNF blocker or MS specialty medication during the period from July 2006 through June 2008. Prescription abandonment was assessed among continuously enrolled members newly initiating TNF blocker or MS therapy. Prescription abandonment was defined as reversal of the adjudicated claim with no evidence of a subsequent additional adjudicated paid claim in the ensuing 90 days. Separate analyses for MS and TNF blocker therapy were performed to assess the association between member OOP expense and abandonment rate using the Cochran-Armitage test for trend and multivariate logistic regression. Members were placed into 1 of the 7 following OOP expense groups per claim: $0-$100, $101-$150, $151-$200, $201-$250, $251-$350, $351-$500, or more than $500. The association of MS or TNF blocker abandonment rate with OOP expense was tested with logistic regression models using the $0-$100 OOP as the reference group and adjusting for age, gender, formulary status, ZIP code-level income and education, earliest specialty medication claim, and methotrexate use for the TNF blocker analysis.
RESULTS:
Of 2,791 members presenting a prescription to newly initiate high-cost MS therapy, 1,985 (71.1%) of the claims were for a 1-month supply with most of the remainder for a 3-month supply; 2,303 (82.5%) had an OOP expense of $0-$100, and 5.4% had an OOP expense greater than $500. The abandonment rate increased as OOP increased (test for trend, P less than 0.001). Members with an OOP expense of $100 or less had an abandonment rate of 5.7%. Among members in all OOP expense groups greater than $200, the abandonment rate was significantly higher, with more than 1 in 4 members abandoning their MS claims (P less than 0.001). In the multivariate logistic regression analysis, the abandonment rate became significantly higher at OOP expenses of $201 to $250 compared with an OOP expense of $100 or less (odds ratio [OR]=7.3, 95% confidence interval [CI]=3.3-16.2). The odds ratios ranged between 6.1 and 7.3 for OOP expense groups greater than $200. Of 7,313 members presenting a prescription to newly initiate TNF blocker therapy, 5,809 (79.4%) of claims were for a 1-month supply with most of the remainder for a 3-month supply; 6,123 (83.7%) had an OOP expense of $0-$100 and 5.7% had an OOP expense greater than $500. The abandonment rate increased as OOP expense increased (test for trend, P less than 0.001). In the multivariate logistic regression analysis, the TNF blocker medication abandonment rate was significantly higher for all OOP expense groups greater than $100, with abandonment odds ratios of 2.3 to 4.4 for OOP expense between $101 and $500 compared with OOP expense of $0-$100. The odds of abandonment at OOP expense of greater than $500 were 7-fold higher